32

During my fourth year as a postdoctoral researcher in computer science at my university, a position for an assistant professor opened up and was available to be filled by one of the local postdocs. We were only two candidates for the position and both completed our PhDs in the same year. However, during the interview process, the selection committee mentioned that my citation count (around 200) and h-index (8) were significantly lower compared to the other candidate.

In the past, I had worked with an exceptional professor for my PhD who emphasized the importance of producing high-quality, original research papers that could only be published in top-tier venues, even if it took a long time to get them published. I have continued to follow this approach in my research, focusing on developing novel ideas and publishing them in prestigious journals and highly selective conferences.

In contrast, my colleague has co-authored a large number of papers with numerous people, but a majority of these papers are surveys or discuss the challenges and issues surrounding certain topics. As a result, my colleague has garnered around 3000 citations and an h-index above 20.

I did not anticipate that the selection committee would necessarily favour me over my colleague, but I did expect to be seen as a viable candidate and for the selection process to be based on small, distinguishing factors. However, based on their criteria, it seems that I have already been eliminated from consideration before the committee has even reviewed my resume.

I am wondering if the strategy that I (and my PhD supervisor) have employed for research and publishing is flawed. Should I start writing survey papers, which have a higher likelihood of being published and cited, in order to improve my citation count and h-index? Alternatively, should I prioritize getting my name listed on as many papers as possible, even if it means compromising my contribution and the quality or originality of my research?

I want to clarify that I did not mean to imply that I am better than my colleague or that I believe myself to be superior. I was simply comparing a specific aspect of our research, the h-index, which played a role in the selection process. I explained the reason for the difference in our h-indices, which was mainly due to the types of publications we have each produced. I have a great deal of respect for my colleague and consider him to be highly competent. I would have been just as pleased if he had been selected, but not because of the h-index criterion. We are colleagues and I will continue to support him. However, my research philosophy differs from his, and I was wondering if this would hinder my progress.

1
  • Some answers-in-comments and other discussion has been moved to chat.
    – cag51
    Jan 5 at 1:05

5 Answers 5

32

There are two separate points here.

First:

However, based on their criteria, it seems that I have already been eliminated from consideration before the committee has even reviewed my resume.

It may well be that, irrespective of h-index etc, the committee had made their mind and the h-index is just a convenient excuse. Choosing a candidate depends on multiple factors, many of which are extremely subjective or self-serving.

Next, ask yourself: do you really want to spend the next 35-40 years working in a place where “developing novel ideas and publishing them in prestigious journals and highly selective conferences” is not a priority? By your own admission it seems you would be a bit of a misfit at that place.

You should think thrice about taking a job at a place where you are unlikely to thrive: this could lead to 35-40 years of misery, and could stunt your research development. Of course it is possible with a good network of collaborators to still prosper in such an environment by delicately balancing sheer productivity and visibility, but it seems your current unit has revealed their cards and what they want is not really what you want.

18

First of, realize that you are currently drawing a lot of conclusions from a single data point. In this one specific job search the committee valued the number of publications and h-index higher than the quality of publications. This is not enough data to determine that your (or your supervisor's) strategy is flawed. Moreover, there is no "strategy" that guarantees success at any specific hiring - different hirings are done by different people, who value different things, and nobody (not your advisor nor anybody else) can tell you with certainty what exactly a specific committee will look for.

That said, in practice what candidates usually need to be successful is a healthy mix of number of publications, strong publications (as proxied by "competitive venues"), and a reasonable citation count for your field and career age (what is "reasonable" depends a lot on the field, but I have to say in applied CS an h-index of 8 would indeed be seen as rather low after 4 years of postdoc). Nobody gets hired onto a good position by publishing 50 borderline-spam papers, but today you will also find it difficult to get hired if you publish substantially less than your colleagues (even if those papers that you do publish are all very strong). How precisely the different dimensions will be weighted will be up to the committee.


One word of warning, not really specifically for OP since I don't know your CV, but generally for those that rely on "quality" to demonstrate research excellence. Doing less but "excellent" research is not a good strategy if that excellence is not easily visible or objectifiable. If you plan on banking on excellence, these excellent papers (a) need to objectively be very strong (it's not sufficient if you find your work much better than the work of competitors, the wider field needs to see it in the same way!), and (b) they need to appear at absolute top venues and garner visible interest (citations may be one way to achieve this, but public discourse etc. could be another). I have met candidates who told me that they don't understand why their really strong papers don't count for more, but upon reviewing their CV I found that all of this work was either unpublished or published at mediocre venues and was effectively uncited. At this point you are banking on a hiring committee detailedly reviewing your papers to determine that they are actually very good, and with dozens (or, in some cases, hundreds) of applications this is a very, very long stretch.

7
  • 4
    What is a reasonable H-index after in applied CS after 4 years of postdoc?
    – Meistro
    Jan 3 at 12:51
  • 2
    I don't know, in my field I would typically expect something between 10 - 15 for an average hireable candidate after a few years of postdoc. But that doesn't mean that we never hire people with 8 (or 20+, for that matter). I think you are looking for absolute metrics that simply don't exist - if you have 8 and your competitor has 20, and the committee cares about that stuff, it's not great for you. Next time you may have 8 and your competitor has 10, and the difference is basically noise.
    – xLeitix
    Jan 3 at 13:20
  • 4
    "find a balance between working on high-quality papers and increasing my citation count and h-index" That's in a nutshell of what I tell my mentees, yes. "Not caring" about anything is generally a risky strategy, since you will meet people that do care about that. Doesn't matter whether that's citations, funding, teaching, service, etc. etc. You can't make everybody happy and excited, but it's good to try to be at least around the average in everything (and very good in some selected things).
    – xLeitix
    Jan 3 at 14:32
  • 3
    Saw this here: "Be aware that many EU institutions are signatories of DORA (sfdora.org) and as such pledge not to use raw numeric metrics in hiring decisions".
    – apg
    Jan 3 at 22:52
  • 2
    @apkg I wonder at how many of these places this "pledge" has actually changed their decision making, and how often it is pure white-washing / marketing.
    – xLeitix
    Jan 5 at 8:09
7

One major realization for me was that high-quality work is good at communication, not just obtaining great results. You may do Nobel prize-worthy work only for it to be ignored and forgotten for decades, if not eternity, if you fail to communicate it well.

We do not have a particularly good way of predicting a candidate's performance. Evaluating a body of work is also hard, especially when it comes to recent papers. But if a candidate has one citation on a paper published 5 years ago in a fast-moving field, this was not an impactful publication, most likely. If they were way "ahead of their time", the communication part of the job was done poorly.

Management boils down to running on a treadmill nonstop: you produce results to acquire more resources to produce better results and acquire even more resources and... Well, if you put too little effort in resource acquisition, it is hard to get very far. Pragmatically, a big part of professorship is enabling others to do great work and attracting funding. Producing lots of low-quality papers is unnecessary; but it is pretty unrealistic to survive without chest thumping altogether - and you won't get higher up the ladder. Like xLeitix says - if your h-index is 8, but all these papers have 100+ citations, you still have a strong case. If they all have about 10, why are you convincing the committee they are absolutely outstanding while you should have been impressing your peers in a broader scientific world instead?

Similarly, a part of the perception here is an inability to turn results into something tangible. You may perceive it as salami publishing, but the powers that be are already planning the resource allocation for the upcoming year. If you want a slice of the pie, you better keep reporting in.

It is what it is. Do not think of promoting your work as something inherently at odds with producing good research - it is another part of the job, it takes time and energy, a lot of both, in fact, but you just fail to get through otherwise. While your progress is apparent to you, it is not the case for other people: to appreciate it fully, they would have to do a lot of work that you do, and that is just infeasible most of the time. This outreach does not necessarily have to be achieved by the means of publishing lots of survey papers, but you still have to convince people your research is useful to them. And by the time you are in front of a committee, it is a bit too late for that.

2
  • I added a couple of edits. But one sentence needs to be rewritten by you yourself as I can make no sense of it: Evaluating past results is also hard, especially regarding fresh papers. Also, can you please clarify how OP might "communicate" his papers better. I take it you do not mean writing them more clearly.
    – Trunk
    Jan 3 at 17:10
  • @Trunk Thank you; I had to revert an edit (since I was speaking about management in the most general sense and not specifically the research management). I have amended that sentence; hopefully, it makes more sense now.
    – Lodinn
    Jan 3 at 17:21
3

You suggest that you did 4 years of postdoc already at that university. This should mean that you already have a good idea of what that Department's priorities are, educationally and research-wise. It should also mean that you have a good idea of the character of the senior staff.

I think that you should reflect on the first point made by ZeroTheHero, i.e that your research's formal metrics are simply a plausible pretext for excluding you.

Beyond research, we can't speculate as you say nothing about other aspects of the job as discussed at interview. Academic purists will find it hard to argue with the advice of your PhD supervisor. But if you hold with this ethos then you have to seek out your own kind in academia. Now jobs are hard to come by anyway, so searching only in known ethos-congenial institutions is unwise.

I was a bit intrigued at your rival's willingness to submit review/summary papers: on the face of it neither of you are senior enough to be doing this. (I also thought that this type of paper had to be requested by senior conference organizers and was not something one could just voluntarily offer to them: we learn something new every day.) No doubt it was done in part for boosting the authors' profiles and giving them a high-visibility platform on which to cite their own papers: these are the margins involved in the growing one's formal metrics of publications.

You need to sharpen your academic elbows a bit or you will continue to be nudged away from the feed-trough by fatter beasts. Maybe take a cold look at your research work to date and see if you can find something like an industrially interesting application from this that might attract funding directly to you. This may involve collaboration of course. But if you can attract funding, the formal metrics may no longer matter much when you are being interviewed: your attraction to a hiring department will be more direct.

3

I believe you need to reevaluate both your and colleague's papers/research output.

Starting with the most obvious, his top 20 papers got 20 citations each. Dividing your 200 citations among at least 8 papers means your average paper got about the same number of citations as his minimum, yet you have half of his good papers. If your papers are really that good, why does nobody bother to cite?

Perhaps, what you are missing is marketing of your papers. How to do it is another matter where I unfortunately cannot help much. Maybe all you are missing are better (more attractive) titles. Or perhaps what you need to do is rethink your strategy - one possibility is having more decent-but-not-great papers. This approach has two benefits over your current approach of "only the best" (with the obvious cost of spending time on that). One benefit is that people can cite any (or even more than one) of your papers, boosting your total citations and other bibliometric scores. Instant improvement to meaningful score. Second and perhaps more important benefit is that someone will read your ok-ish paper - and see that you cite your own great paper, making them read it - again, leading people to read and cite your best papers.

Second problem I have with your characterization of being singled out - you are effectively saying colleague's papers are meh as they are just surveys or only show issues/challenges - but some of the most important papers in many fields are those that show currently mainstream strategy is fundamentally flawed and can't work. Yes, review papers get more citations often for less work and that stings and stinks. I would hope that committee does read few best papers of top candidates to account for that. If they did, perhaps they weren't as impressed by your top papers as you are.

Here, the best way forward would be to find his top few publications, and select your top few publications, and ask someone for his informal opinion. Perhaps your supervisor, though it would be better if you could find someone impartial. It is always the best to judge papers with fresh and unbiased set of eyes. Almost everyone thinks his own research is better and more important than it really is (as judged by the whole society).

7
  • 2
    1) I want to make it clear that I did not intend to present myself as superior to my colleague. I respect his work and do not think poorly of it or that his papers are meh. I don't know how you got that from my question. 2), the number of citations a paper receives does not necessarily reflect its quality. It is well-known that some papers are more likely to be cited due to their visibility on search engines or other bibliographic indexes.
    – Meistro
    Jan 3 at 19:01
  • 1
    3) I did not make any statements about the number of citations my papers have received. I simply pointed out that survey papers tend to be cited more frequently than other types of papers, which is a logical observation. 4)( some of) the papers are good because they are accepted at venues where the acceptance rate below 10%. While I may have concerns about the review process in academia, I still believe that acceptance at such venues is a credible indication of a paper's quality.
    – Meistro
    Jan 3 at 19:01
  • Please elaborate on "marketing of [OP's] papers".
    – Trunk
    Jan 3 at 19:10
  • 1
    @Trunk One way of marketing is having more papers that also serve to boost your own visibility - people will notice another of your papers and cite the really important one. Another might be posting on Linkedin/Researchgate/FB/... But in general I have no idea how to approach this marketing thing so I didn't elaborate more - I am primarily observing that if he isn't getting citations it might be that despite quality, papers are not noticed enough. Jan 3 at 22:06
  • 1
    @Zizy Archer Yet self-promoting papers are the essence of the pulp production that OP disdains. Why do staff citation metrics matter so much? Presumably because they elevate the department's ranking and this makes research council funding easier. If OP and his favorite collaborators work together on one or two more applied research projects this may offer his future academic employer a more tangible financial bounty in hiring him. And with significant industry finding attached to him the path to a full-tenure position is clearer.
    – Trunk
    Jan 3 at 23:04

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .